Transferring data from non Unicode R3 system to Unicode system via PI

[Former Member]

Hello Experts,

I am working on transfering data from a non unicode R/3 system to Unicode system via PI/XI and viase versa. I have the following queries:

Non-Unicode to Unicode:

What happens if the Identified language key is not consistent with the text value that exists for eg: if English text in non-unicode system has Spanish characters within..

Does PI have all parameters required to recognize the MDMP details of the source/target system (Business/Technical definition) and do the necessary conversion accordingly

Unicode to non-Unicode:

The understanding is that Unicode can contain more characters than a non-unicode (for a specific language). How do we handle a scenario where the additional characters (that are in unicode version & does not exist in non-unicode one) are sent from Unicode to non-unicode based system.

Your expert views will be highly appreciated.

Thanks,

Yash

[aashish_sinha]

Hi,

The XI Integration Server is installed and operates on top of SAP Web Application Server 6.20 running in Unicode mode. Therefore, a Unicode payload is allowed for message transport and mapping, and messages with Unicode payload can be exchanged with other systems also running in Unicode mode.

The XI add-on for SAP business systems, however, runs with the business system on the respective platform, be it Unicode or not.

Therefore, when interfacing with SAP business systems running in a single, non-blended code page (single-byte or double-byte), the IDoc or RFC adapter is used for code page conversion to and from Unicode for message payloads. When interfacing with non-SAP business systems, the Plain HTTP adapter or the adapters provided by the Adapter Engine are used for this purpose.

Using the XI message processing facilities does not necessarily enable successful communication between business systems with different code pages. It only enables business systems with any code page to successfully communicate with the Integration Server. This means that running XI in Unicode mode does not solve any character conversion problems (for example, non-compatible code pages and resulting information loss) that would also appear in a direct connection of two systems without XI.

To sum up, all this means that XI on Unicode delivers additional advantages but does not solve all code page mapping problems. The additional advantages are:

- Communication with Unicode business systems for IDocs and RFC calls supported.

- Communication between groups of business systems with transferable code pages supported.

- Communication between business systems with identical code pages still ensured.

The integration with systems running with blended or multiple code-pages (MDMP) will have to be done on individual project basis.

Please check sap note 548016 for Conversion from non-Unicode to Unicode; system copy of a Unicode system and SAP Note 508854 for How to use Unicode mode of SAP frontend and the restrictions of it

Hope this will help you.

Regards

Aashish Sinha

PS : reward points if helpful

[Former Member]

Hi Aashish,

Thnaks for your quick response. I have gone thru the notes and they seems to be very generic. I have the following queries to be specific:

Non-Unicode(r3) to Unicode()MDM:

How do we plan to handle multiple language issues, for those supported in R/3 & not in MDM

all non-unicode characters for a specific language will be available in Unicode version of the language

Does PI have all parameters required to recognize the MDMP details of the source/target system (Business/Technical definition) and do the necessary conversion accordingly

What happens if the Identified language key is not consistent with the text value that exists for eg: if English text in non-unicode system has Spanish characters within..

Unicode to non-Unicode:

How do we plan to handle multiple language issues, for those supported in MDM & not in R/3 (or maybe any other mainframe system/..)

Which system (PI/ MDM) will have the information on the languages that can be supported in the target system?

The understanding is that Unicode can contain more characters than a non-unicode (for a specific language). How do we handle a scenario where the additional characters (that are in unicode version & does not exist in non-unicode one) are sent from Unicode to non-unicode based system.

Hope to got my concern. Your reponse will be highly appreciated.

Thanks,

Yash